Skip to content

Pull requests: NVIDIA/Model-Optimizer

New pull request New
171 Open 1,038 Closed
171 Open 1,038 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Calibrate vLLM non-gated MoE through Triton fakequant
#1572 opened May 30, 2026 by mxinO Contributor • Draft
4
feat: Layerwise calibration: nested config + QDQ-from-prev-layer flag + checkpoint I/O knobs
#1571 opened May 29, 2026 by Fridah-nv Contributor Loading…
32
Update Qwen3-8B EAGLE3 example to online training pipeline
#1570 opened May 29, 2026 by cjluo-nv Collaborator • Draft
2
[6078291][OMNIML-3716] Add ViT FP8/NVFP4 recipes + Torch-TRT example, wire softmax_quantizer in _QuantAttention
#1569 opened May 29, 2026 by ajrasane Contributor Loading…
9
[OMNIML-4788] specdec_bench/Qwen3.5-4B: throughput_32k benchmark + S3 upload step
#1564 opened May 28, 2026 by ChenhanYu Collaborator Loading…
10
FSDP2 calibration with hf_ptq.py
#1563 opened May 28, 2026 by sugunav14 Contributor • Draft
8
Autoquant and GPTQ in support in Megatron-Core [OMNIML-3151]
#1562 opened May 28, 2026 by jenchen13 Contributor Loading…
5
[OMNIML-3994] Make sure all weight quantizers have _amax
#1560 opened May 28, 2026 by sychen52 Contributor Loading…
@sychen52
19
[5924759] Fix fp16 ONNX INT8 entropy calibration on numpy >= 2.0
#1558 opened May 28, 2026 by ajrasane Contributor Loading…
7
fix(quant): sync NVFP4StaticQuantizer global_amax across TP and EP
#1553 opened May 28, 2026 by cjluo-nv Collaborator • Draft
4
Add Codex wrappers for Claude skills
#1551 opened May 28, 2026 by meenchen Contributor Loading…
@meenchen
7
WIP Support per expert amax in TEGroupedMLP
#1550 opened May 27, 2026 by jenchen13 Contributor Loading…
5
Add modelopt agent protocols including shared schemas with bigpareto and quant_flow
#1549 opened May 27, 2026 by Edwardf0t1 Contributor • Draft
3
docs(launcher): scope CLAUDE.md "new model config" section to Megatron-LM PTQ
#1547 opened May 26, 2026 by h-guo18 Contributor Loading…
2
Scratch: NVFP4 activation input_scale calibration study
#1545 opened May 26, 2026 by cjluo-nv Collaborator • Draft
3
[6/n] Replace skip-softmax calibration formula
#1541 opened May 25, 2026 by kaix-nv Contributor Loading…
10
Fix --trust_calibration_data being mutually exclusive with calibration data paths
#1540 opened May 24, 2026 by adityasingh2400 Loading…
4
Add parallel conv fusion quantization pass
#1538 opened May 23, 2026 by NotUr77 Loading…
2
fix(export): align offloaded modules before pre-quant-scale fuse (#795) bug Something isn't working
#1530 opened May 21, 2026 by realAsma Contributor Loading…
4
[Tests]: Precommit Check for Spec-Dec Recipes
#1527 opened May 21, 2026 by h-guo18 Contributor Loading…
6
[Recipes]: Add specdec per model recipes
#1526 opened May 21, 2026 by h-guo18 Contributor • Draft
@h-guo18
4
Drive hf_ptq qformat choices from preset YAMLs (remove hardcoded CLI quant configs)
#1525 opened May 21, 2026 by shengliangxu Collaborator Loading…
8
Add YAML based AutoQuantize recipe (currently only CLI is supported)
#1523 opened May 21, 2026 by juhi10071998 Contributor Loading…
4 of 6 tasks
25
[fix] Fix NVFP4 AWQ GQA prequant fusion
#1520 opened May 19, 2026 by ShawRong Loading…
5
fix(quantization): accept QuantizeAlgorithmConfig in get_modelike_from_algo_cfg
#1514 opened May 18, 2026 by javierdejesusda Contributor Loading…
1
7
ProTip! no:milestone will show everything without a milestone.